This paper provides an evidence-based risk-benefit assessment of acamprosate and naltrexone in the treatment of alcohol dependence. A risk-benefit assessment is based on the premise that the choice of treatment depends on a number of factors, notably the adverse event profile and efficacy. An evidence-based approach attempts to operationalize how such risk-benefit assessments are made to inform physician choices. This approach involves a systematic assessment of all published double-blind, placebo-controlled trials. Based on this review, we conclude acamprosate and naltrexone are both useful in the treatment of alcohol dependence. However, the two drugs act in different ways in the brain, and their clinical profiles are different. Treatment effects seem to be more reliable for acamprosate, and this drug is better tolerated. The safety of the two drugs in combination has been supported by two independent double-blind studies, and combination treatment may offer an advantage for some patients.